D3M Core is in the process of updating Python from 3.6 to 3.7, 3.8, and eventually 3.9. As part of the process, we'll need to update the pandas version to at least 1.1.5. Could you verify that the metalearn package will work with pandas 1.1.5 and update accordingly? We would prefer if you could specify the pandas dependency to be >= 1.1.5 so we can avoid this issue in the future.
Hi @bhavints! I understand the issue, but I am not currently available to work on this. Updating the pandas version would not take too much time, but submitting primitives has proven to be a difficult process.
Hi all,
D3M Core is in the process of updating Python from 3.6 to 3.7, 3.8, and eventually 3.9. As part of the process, we'll need to update the pandas version to at least 1.1.5. Could you verify that the metalearn package will work with pandas 1.1.5 and update accordingly? We would prefer if you could specify the pandas dependency to be >= 1.1.5 so we can avoid this issue in the future.
D3M images are failing to build due to pandas version issue: https://gitlab.com/datadrivendiscovery/images/-/jobs/1667009304.
@bjschoenfeld